Channel Length Modulation in MOS Circuit Design

I've had some long-standing confusion in this subject and decided to
write a paper to attempt to clarify things for myself. I'd be interested
to see if others think I'm completely off-track or think my points are
valid.

  http://galois.digitalsignallabs.com/channellengthmodulation.pdf

PS: I know there's a lot of sharp, eclectic engineers on comp.dsp, hence
my cross-post there.
-- 
%  Randy Yates                  % "And all that I can do
%% Fuquay-Varina, NC            %  is say I'm sorry, 
%%% 919-577-9882                %  that's the way it goes..."
%%%% <yates@ieee.org>           % Getting To The Point', *Balance of Power*, ELO
http://www.digitalsignallabs.com

Hi Randy,

I went quickly through your pdf this WE, well done ! Well I have any
pretension of arguing your equations but I think you know what you're
talking about. Anyway, there is one point that would not suit one of
my university teachers. He was always getting very upset and angry
when seeing the word 'Early' in a MOS-related document. In fact, the
similarity in effect between the Early effects in bipolar and channel-
length modulation in MOS makes people talking about 'Early' in the MOS
world as too. My teacher didn't like it at all ... ;-) But don't worry
about it, many people, including myself, are using this mixture.
There is a paper on the IEEE that can help you in better understanding
and/or cross-checking your results:

An exponential model of channel-length modulation applied towards
floating-gate circuits
http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/7554/20575/00951395.pdf?arnumber=951395

Have fun !

Riad.